CoursewareHub, which is developed to use Jupyter Notebook in lectures and exercises, can acquire learning logs related to students’ cell execution, and is being used to improve classes and support classes based on the results of the log analysis. The authors have proposed a cloud-based log analysis environment using Elasticsearch, Fluentd and Kibana, and have extended the scope of application to network devices and LMS. In this paper, we aim to provide an environment for log analysis that enables correlation analysis of multiple logs by further expanding the scope of application of this environment. As an example, we applied CoursewareHub logs to this environment and constructed a dashboard that enables teachers conducting classes to grasp the status of students by analysing CoursewareHub logs in real time.

Internet of Things (IoT) is widely used as a fundamental technology for realizing various services. An IoT-based service system comprising cloud servers and many IoT devices connected via networks may be risky owing to the possibility of the entire system being be affected by cyberattacks on an IoT device. Moreover, new software vulnerabilities are frequently reported. From the perspective of security, IoT device software must be reliable and resilient. Consequently, a secure software update mechanism for IoT, assurance of software reliability, and mitigation mechanisms are required. Overall, this study proposed a zero-trust over-the-air (ZT-OTA) update framework for the reliable and resilient software update management of IoT devices via OTA software updates from remote locations. The ZT-OTA update framework provides a strict software version-management mechanism that enhances the security of software updates. Further, the proposed framework collaborates with a Software Assessment Service (SAS) as an authorized third-party assessment organization and deploys reliable software approved by the service to IoT devices. Upon discovering a software vulnerability following the deployment of the software, the SAS proactively revokes the reliability of the software by revoking its certificates for signing software codes and the vulnerability assessment result. Subsequently, the ZT-OTA server notifies developers and users of IoT devices to arrange for new software that must be fixed and automatically acts on behalf of IoT devices that must be retired. This study introduces the detailed design of the ZT-OTA update framework and describes to demonstrate its feasibility.

In this study, we aimed to construct a system to perform statistical analyses for educational improvement while preserving privacy. Various probabilistic statistical algorithms have been developed for data agitation. However, one of the major challenges in using these algorithms is determining appropriate parameters. To address this issue, our previous research proposed the “α-criterion” as a criterion that must be satisfied by appropriate parameters. This paper outlines a system that calculates the parameters that satisfy the α-criterion while performing real-time data agitation on the given aggregate data. To verify that our implementation actually works, we show a use case with sample data on kibana-Elastic log analysis provided by Elastic N. V.

We propose a criterion of reasonable parameters for an algorithm that aggregates target data with anonymized data for safe data analysis in online educational systems, called learning management systems (LMSs). We also statistically investigate parameters that can satisfy the proposed criteria using an anonymization algorithm and real large-scale data. We use an open dataset containing one year's worth of product review data due to the difficulty of collecting LMS data large enough for the evaluation. Furthermore, we discuss an approach to address cases where no parameter satisfies the proposed criteria.

Demands for Internet of Things (IoT) platforms are increasing along with the expansion of data-driven sciences, and academic network infrastructure such as national research and education networks (NRENs) are now being encouraged to aggressively support diverse IoT research projects. In Japan, the National Institute of Informatics (NII) operates an NREN called "SINET5," which is an academic backbone network linking more than 900 universities and research institutions. In addition to SINET5's high-speed 100 Gbps backbone network, SINET5 also provides a mobile network called "Mobile SINET" as IoT infrastructure. To provide crucial support to IoT research projects and to facilitate the development and deployment of IoT applications, we present herein our experience with our software library, "SINETStream." In this paper, we provide an overview of SINETStream and discuss lessons learned from examples of application deployment over Mobile SINET. SINETStream provides a common and simple application program interface (API) for various message brokers, security functions, and performance tuning support features. These functions improve the portability of applications and help application developers to remove hindrances to the development of secure and efficient IoT applications. The experimental results provided herein show that application developers can use SINETStream functions within a reasonable overhead. We also show how the combination of SINETStream and Mobile SINET enables users to develop and deploy highly confidential and efficient IoT applications.

As learning management systems (LMSs) become increasingly popular in many universities, large scale Moodle systems, such as cross-university LMSs, have been implemented, and they accumlate a large amount of LMS log data. To handle such a large amount of data, it is effective to use a cloud computing environment. However, because LMS log data contain much personal information, it is often necessary to anonymise or pseudonymise the data to use them in cloud computing environment. In this study, we implemented a log analysis environment that combines an organisation’s internal server and a cloud computing environment by using the GakuNin-Cloud on-demand configuration service provided by the National Institute of Informatics (NII). In this environment, LMS logs are stored in the cloud with pseudonymisation, and the pseudonymisation is automatically removed by a re-identification proxy when the logs need to be analysed. We applied the environment to the Moodle system at Gunma University. We also applied the environment to an example Moodle created by the GakuNin Cloud on-demand configuration service.

As the foundation of intelligent algorithms and applications, data collection from the real world faces the problem that there is serious data degradation under various complex environments. As a typical situation, the visual degradation of images under different weather conditions only can be utilized after arduous image noise removal by application developers previously. To overcome the challenges, previous approaches cannot handle with comprehensive situations. In this paper, we will briefly describe an adaptive image noise removal tool, which can classify multiple weather conditions and enhance image quality with optimized algorithms. Further, we constructed a recognition application using YOLO-3, and validated the effect of our tool through recognition results of real-world images.

In this era of cloud computing, users encounter the challenging task of effectively composing and running their applications on the cloud. By understanding user behavior in constructing applications and interacting with typical cloud infrastructures, cloud managers can develop better systems that improve the users' experience. In this paper, we analyze a large dataset of a Google cluster to characterize the users into distinct groups of similar usage behavior. We used a wide range of measured metrics to model user behavior in composing applications from the perspective of actions around application architecting, capacity planning, and workload type planning and to model user interaction behavior around the session view. The trajectories of users' actions are represented as sequences using categorical and proportional encoding schemes. We used techniques from the sequence analysis paradigm to quantify dissimilarity among users. We employed a robust cluster analysis procedure based on the agglomerative hierarchical methods to optimally classify users into 12 classes. We used a variety of formal indices and visual aids to confirm the quality and stability of the outcomes. By visual inspection, we regrouped the obtained clusters into 5 main groups that reveal interesting insights about the characteristics which underline different groups' utilization behavior.

Linux containers have become very popular these days due to their lightweight nature and portability. Numerous web services are now deployed as clusters of containers. Kubernetes is a popular container management system that enables users to deploy such web services easily, and hence, it facilitates web service migration to the other side of the world. However, since Kubernetes relies on external load balancers provided by cloud providers, it is difficult to use in environments where there are no supported load balancers. This is particularly true for on-premise data centers, or for all but the largest cloud providers. In this paper, we proposed a portable load balancer that was usable in any environment, and hence facilitated web services migration. We implemented a containerized software load balancer that is run by Kubernetes as a part of container cluster, using Linux kernel's Internet Protocol Virtual Server(IPVS). Then we compared the performance of our proposed load balancer with existing iptables Destination Network Address Translation (DNAT) and the Nginx load balancers. During our experiments, we also clarified the importance of two network conditions to derive the best performance: the first was the choice of the overlay network operation mode, and the second was distributing packet processing to multiple cores. The results indicated that our proposed IPVS load balancer improved portability of web services without sacrificing the performance.

High-performance R&E networks and clouds enable the construction of a secure virtual cloud environment over their resources. However, it is not easy for general users to construct such an environment and build an application over the environment, as this would require each user to not only have knowledge about the target application, but also about the cloud APIs and resource optimization. This prompted us to propose a Virtual Cloud Service System (VCSS), which allows general academic users to construct an application-specific virtual cloud environment. The system consists of the Virtual Cloud Provider (VCP) middleware and Application Templates. The VCP provides users with a simple service interface and enables them to construct a virtual cloud. The Application Templates provide Jupyter Notebook-based procedure manuals and a management environment for a virtual cloud constructed by the VCP for typical research and educational applications such as a learning management system. The experimental results show that the VCP can provide a secure and high-performance virtual cloud environment over academic and public clouds and the R&E network between them.

Science computing platforms are changing from traditional on-premises computing platforms to clouds. The style of research publications is also changing with the movement of open science. For example, it is common that researchers in the bioscience research community publish their research papers with the associated research data on the Internet. Furthermore, demands for reproducibility of computational experiments are increasing. A conventional method, or sharing source codes of application programs, is not enough to guarantee the reproducibility of computational experiments. In this paper, we propose overlay cloud architecture for building virtual clouds on cloud platforms. In the proposed architecture, the middleware, Virtual Cloud Provider (VCP), automatically configures computing environments (application programs, operating systems, libraries and binaries) that is required to reproduce computing results by using the Linux container technology. VCP also deploys containers of computing environments on clouds and configures overlay network connecting the deployed containers. A case study which considers DNA sequencing software shows how the VCP can be deployed, and used to reproduce results.

In the era of cloud computing, users encounter the challenging task of effectively composing and running their applications on the cloud. In an attempt to understand user behavior in constructing applications and interacting with typical cloud infrastructures, we analyzed a large utilization dataset of Google cluster. In the present paper, we consider user behavior in composing applications from the perspective of topology, maximum requested computational resources, and workload type. We model user dynamic behavior around the user's session view. Mass-Count disparity metrics are used to investigate the characteristics of underlying statistical models and to characterize users into distinct groups according to their composition and behavioral classes and patterns. The present study reveals interesting insight into the heterogeneous structure of the Google cloud workload.

Resource allocation is an active direction of research that is drawing interest within academic and technological circles. Resource allocation imposes numerous challenges. This is especially true for Inter-Clouds, a recent paradigm for horizontal expansion and integration of disparate and heterogeneous cloud platforms. In an attempt to realize an efficient resource management system, this work-in-progress paper proposes and describes a new multi-layered management framework to address the tasks of virtualized resource control, dynamic resource provisioning, life-cycle management and resource exchange within Inter-Cloud environments.

A parameter survey application (PSA) is a typical application running on high-performance computing (HPC) systems. A PSA consists of a lot of independent tasks with different input parameters that are executed in parallel on different CPU cores. Infrastructure-as-a-Service Cloud (IaaS Cloud) is expected to be used as an HPC infrastructure to run PSAs, and some reports have discussed hybrid execution mechanisms to utilize both local resources and IaaS Clouds. However, users still have security problems in running applications with confidential data on an IaaS Cloud. We propose a hybrid and secure execution mechanism to run PSAs utilizing both local computing resources with a batch scheduler and an IaaS Cloud. The proposed mechanism utilizes both local resources and IaaS Clouds to meet the deadline of user applications. We conducted experiments running a natural language processing application, which uses machine learning to detect abusive language on Internet bulletin board systems. The experimental results showed that the proposed mechanism effectively allocated resources and met the deadlines of the user application.

This paper investigates the scheduling of mixed-parallel applications, which exhibit both task and data parallelism, in advance reservations settings. Both the problem of minimizing application turn-around time and that of meeting a deadline are studied. For each several scheduling algorithms are proposed, some of which borrow ideas from previously published work in non-reservation settings. Algorithms are compared in simulation over a wide range of application and reservation scenarios. The main finding is that schedules computed using the previously published CPA algorithm can be adapted to advance reservation settings, notably resulting in low resource consumption andthus high efficiency.

This paper discusses the impact of the hierarchical master-worker paradigm on performance of an application program, which solves an optimization problem by a parallel branch and bound algorithm on a distributed computing system. The application program, which this paper addresses, solves the BMI Eigenvalue Problem, which is an optimization problem to minimize the greatest eigenvalue of a bilinear matrix function. This paper proposes a parallel branch and bound algorithm to solve the BMI Eigenvalue Problem with the hierarchical master-worker paradigm. The experimental results showed that the conventional algorithm with the master-worker paradigm significantly degraded performance on a Grid test bed, where computing resources were distributed on WAN via a firewall; however, the hierarchical master-worker paradigm sustained good performance.

While there have been several proposals of high-performance global computing systems, scheduling schemes for the systems have not been well investigated. The reason is difficulties of evaluation by large-scale benchmarks with reproducible results. Our Bricks performance evaluation system allows the analysis and comparison of various scheduling schemes in a typical high-performance global computing setting. Bricks can simulate various behaviors of global computing systems, especially the behavior of networks and resource scheduling algorithms. Moreover, Bricks is partitioned into components such that not only can its constituents be replaced to simulate various different system algorithms, but it also allows the incorporation of existing global computing components via its foreign interface. To test the validity of the latter characteristics, we incorporated the NWS (Network Weather Service) system, which monitors and forecasts global computing systems behavior. Experiments were conducted by running NWS under a real environment versus a Bricks-simulated environment, given the observed parameters of the real environment. We observed that Bricks behaved in the same manner as the real environment, and NWS also behaved similarly, making quite comparative forecasts under both environments.

The paper proposes a performance evaluation model for effective job scheduling in global computing systems. The proposed model represents a global computing system by a queueing network, in which servers and networks are represented by queueing systems. Evaluation of the proposed model showed that the model could simulate behavior of an actual global computing system and job scheduling on the system effectively.